3 ms·

I have a project dedicated to this topic: https://github.com/ClickHouse/ClickBench https://github.com/ClickHouse/ClickBench It is important to explain the limi

I have a project dedicated to this topic: https://github.com/ClickHouse/ClickBench https://github.com/ClickHouse/ClickBench It is important to explain the limitations of a benchmark, provide a methodology, and make it reproducible. It also has to be simple enough, otherwise it will not be realistic to include a large number of participants. I'm also collecting all database benchmarks I could find: https://github.com/ClickHouse/ClickHouse/issues/22398 https://github.com/ClickHouse/ClickHouse/issues/22398

leeoniya 3y ago
oh, hey there :) yep. no matter how much bench homework i try to do with my GH projects (uPlot, uFuzzy, uDSV, uTable) to show off their perf, it's impossible to be anywhere near exhaustive. there is simply too much variance in envs, hardware, software versions, etc. i do think it's important that all compared libs perform the same benchmarks though so it's at least apples-to-apples, even if it's just 5 apples out of 1,000 :) it's a lot of work implementing (in good faith) all other competitors' libs.